Thank you all for joining us today for Castle's 2022 Investor Day. Before we begin, I would like to remind you that some of the statements made today will contain forward-looking statements within the meaning of the Private Securities Litigation Reform Act of 1995. Please take a moment to review these statements.
We'll start today by hearing from Derek Maetzold, Castle's Founder, President, and CEO, and he'll discuss our overall corporate strategy. Dr. Brent Moody of Heritage Medical Associates, a thought leader in dermatology and Mohs surgeon, will then discuss two of our skin cancer tests focused on risk stratification, DecisionDx-Melanoma and DecisionDx-SCC.
We'll then have Dr. Matt Goldberg, dermatologist, dermatopathologist, and our head medical director, discuss our diagnostic GEP offering, MyPath Melanoma and DiffDx-Melanoma.
He will then give an overview of our inflammatory skin disease pipeline initiative, which is our test in development to predict therapy response. Following Dr. Goldberg, Dr. Craig Monroe, a very recent addition to the Castle team, gastroenterologist and medical director of our GI franchise, will discuss TissueCypher, our Barrett's esophagus test. Dr. Bob Cook, Senior Vice President of R&D at Castle, will discuss IDgenetix, our PGX test focused on mental health.
I'll then go over our financial outlook, and Derek will give a summary, and we'll open it up for Q&A.
Frank, thank you for the introductions. Good Afternoon, everyone, and thank you for making the decision to participate in our inaugural Investor Day. We hope that you leave this virtual meeting with a deeper understanding of our business plans, our opportunities, our risks, and are in a better position to appreciate the midterm expectations that we, as a management team, have for the profitable growth of Castle Biosciences.
Let me begin by talking a bit about Castle and our philosophy. We are, as you know, a molecular diagnostic testing company which develops proprietary laboratory-developed tests.
Although we have grown significantly since going public in mid-2019, we remain focused on the three statements you see here. Our goal, our mission as a company is to improve the health of our patients through development of our innovative tests.
We have, as a company, which includes each and every one of our employees, remain steadfastly committed to this mission. If we accomplish our mission from a test development standpoint, from a patient care standpoint, and from an overall mission perspective, it is our expectation that we will transform patient management by keeping people first.
That is, the patients we serve, their clinicians, our employees, and you, our investors. As an aside, I do think it's important to acknowledge that there are many jobs, many careers that one can grab a hold of in the U.S. marketplace.
I firmly believe that those of us who migrate towards healthcare have an internal drive, internal clock that says, "At the end of the day, I want my contribution to life to improve the health of patients." From my perspective, we have and continue to attract individuals who call Castle home, who live this vision every day. I struggle to tell you, as a founder and CEO of Castle Biosciences, how rewarding it is to lead our team with the knowledge that we are all moving in the same direction.
That is, to transform disease management by keeping people first and by focusing on the value that we can bring to their disease management decisions. Finally, we believe we have developed a very strong culture here at Castle, and we believe that is our values that assist in nurturing and maintaining our culture.
These values are lived out every day through maintaining a community of trust from a company perspective, celebrating and pursuing excellence in terms of execution and performance, collaborating, operating with a high level of integrity, focusing on innovation, and creating excitement around the changes we are striving for in patient care. We'll unwrap the story of Castle in the next few slides.
The purpose of this introductory slide is to remind you that we, Castle Biosciences, are an innovator in diagnostic risk stratification and therapy-guiding tests for disease states that have high unmet clinical needs. We firmly believe that the traditional approach of developing treatment plans that are limited to using clinical and pathology factors alone is inadequate and can be improved by incorporating personalized genomic information.
Each and every one of our tests supports this belief. Adding molecular knowledge underscoring a patient's disease does make a difference.
From a customer call point perspective, we have tests that guide patient management decisions today in skin cancer management, in uveal or ocular melanoma management, in gastroenterology management, and more recently, in mental health conditions. Now, I'd like to talk about the three strategic guideposts that we consider the foundation of our business.
Think of them as strategic guiding principles that we adhere to in order to keep us on the right track as we evolve, grow, and continue to create value for our stakeholders.
First, we believe in attracting and cultivating exceptional employees. We hire the right people and keep the right people by Castle's commitment to doing the right thing for employees and nurturing our thriving culture.
It's our belief that attracting and retaining superior talent gives us a competitive edge and that developing this talent is of tremendous benefit to both Castle and our talent pool. That is why we invest in establishing a strong collaborative culture that encourages engagement, and we're proud to have been recognized by our employees in some noteworthy top workplace awards in Arizona, Texas, and even nationwide.
We feel gratified by this and think it is reflective of a highly engaged, motivated, and loyal workforce. Second, we look to continuously evolve and improve.
We are leading the field in molecular diagnostics by challenging the status quo with deep scientific expertise, unique value insight, and robust data development. Serving unmet clinical needs through innovative tests is how Castle got its start, and we've been using novel new approaches to help guide clinical treatment plan decisions ever since.
Simply put, we believe the status quo or traditional approach can be vastly improved upon to attain better outcomes, and this belief has been one of our guiding principles since our inception. Finally, we are customer and solution-centric. We value best-in-class customer experience at all points along the testing journey, and we leverage multiple solutions for a single customer to provide a single source of high-quality molecular diagnostic test solutions.
From a diagnostic support to risk stratification, we have a skin cancer product line that comprises a broad range in the patient care continuum, and we have continued to leverage our resources to branch out into the therapy response area. For example, with our inflammatory skin disease pipeline test and other programs focused on the dermatology call point. Though we expect these to launch by 2025, it demonstrates how we are positioning ourselves to be a more cust...
Comprehensive customer-focused company as we grow. Let's dig a bit deeper into these strategic guideposts and provide you with a deeper understanding of what we focus on. Let's talk first about our exceptional employees, and here you can see some of the highlights and examples of how we try to maximize employee potential and promote an open and engaging atmosphere at Castle.
Through our leadership and professional development programs, we equip our employees to succeed and help Castle continue to flourish.
We want our employees to have the knowledge and expertise needed to succeed and advance their careers, as well as the confidence to solve today's challenges. That's why we invest in their development and professional education, as it's a clear win-win for everyone and goes a long way towards building a better future for Castle and its workforce.
Our investment in our employees has clearly helped us retain talent, with an overall retention rate of 90% in 2021, with just a 2% regrettable turnover rate in that same year. We also have had a strong level of employee engagement, which does go hand in hand, of course, with high levels of retention. We value the unique perspective our employees bring to the organization and encourage open channels of communication.
We build channels for employees to provide feedback and propose solutions, regularly perform engagement surveys for all recent hires, and instituted a voluntary thank you card program that enables any employee to acknowledge other team members for a job well done. We believe this must be resonating with our employees as our 2021 engagement score of 83% was well above the overall healthcare benchmark of 66%.
In 2022, our engagement score was 81% compared with the overall healthcare benchmark of only 53%. Castle also offers defined pathways for advancement, enabling our employees to more fully realize their potential and take on greater levels of responsibilities as their skills develop. Again, we understand the importance of supporting and investing in our staff as their growth is closely tied to Castle's own growth.
Our second strategic guidepost, continuous evolution and improvement, is rooted in one of the core tenets that Castle was founded upon.
That is that traditional methods to determine a patient's cancer prognosis can be improved by applying newly emerging techniques in molecular biology. We continue to focus on transforming disease management in areas of high unmet clinical need through the development and marketing of innovative diagnostic and prognostic tests that guide patient care.
One noteworthy example of this would be the evolution of our DecisionDx-Melanoma test. Our approach has been to identify, develop, and validate genomic classification signatures that are added to the available clinical and pathologic factors.
We do this by demonstrating consistent test performance that is also statistically independent from those clinical pathologic factors, so if the clinician and patient know that adding our genomic classifier information adds value to their clinical decisions. However, we aren't done then.
We do continuously evaluate as to whether we can make our test more actionable. This can come, for instance, in the case of our DecisionDx-Melanoma, where we expanded the use of our test from risk stratification management decisions related to distant metastasis to more recently guiding some of the biopsy surgical decisions in those same patients. We also attack continuous improvement in other ways.
For example, also with the case of DecisionDx-Melanoma, we evaluated the value of incorporating clinical and pathologic factors into our molecular 31-gene expression profile classifier test to see if we could improve the accuracy of our risk stratification calls. And we succeeded by the way. Using deep learning techniques, we developed an AI-based neural network algorithm that we call the i31-GEP test, with i standing for integrated.
This improved version of our DecisionDx-Melanoma test has been shown to provide a more precise, a more personalized risk prediction over both our original 31-GEP test as well as clinical and pathologic factors alone. What a great example of continuous improvement. Another good example would be one of advancing or improving robust data development.
Specifically, our collaboration with the National Cancer Institute to link our DecisionDx-Melanoma testing data from the SEER Cancer Registries Program.
The initial output of this collaboration was a significant development as it yielded new data in real-world clinical practice, demonstrating that patients who were clinically tested with DecisionDx-Melanoma, whose clinicians have the benefit of a patient's clinical and pathologic factors traditionally used to assess risk, and our test result showed a 27% improvement in melanoma-specific survival compared to patients with only traditional clinical pathologic factors available to them to determine their treatment and follow-up plans.
These results have helped to promote and differentiate our tests and have driven greater adoption and test volume since being publicized earlier this year. Our pipeline initiatives are also a great example of leveraging our expertise in dermatology to expand into other prospective areas of growth.
Specifically, we are developing a genomic test aimed at predicting therapy response in patients with inflammatory skin conditions. While we are still in the development and validation process, we do expect to have a clinically available test by the end of 2025. If successful, this new test has the potential to add approximately $1.9 billion to our current US TAM. Last year, we signed an agreement with Modernizing Medicine to establish an interface with their electronic health record system, EMA.
This interface is expected to streamline the ordering process and allow clinicians to order Castle's full suite of skin cancer tests from directly within the patient's medical record and also receive Castle's test results from within that same system.
This is clearly a big improvement in both convenience and efficiency for our customers as well as Castle, and it straddles the line with our third strategic guidepost. Being customer and solution centric is one of our foundational focus areas as we believe in delivering a great customer experience and leveraging multiple solutions to provide customers a single convenient source for multiple molecular diagnostic tests.
You see this most prominently today in our dermatology franchise, in which we offer tests at multiple stages of a patient journey. For example, MyPath Melanoma and DiffDx-Melanoma in diagnostic support, DecisionDx-Melanoma in risk stratification support, DecisionDx-SCC, also in risk stratification support, and our pipeline test for inflammatory skin disease therapy response. All of these tests align with our initial dermatologic call point.
That being said, our Cernostics and AltheaDx acquisitions were carefully considered for their growth potential and attractiveness as commercial call points as well, and we hope to build on their current product line to offer a greater breadth of tests in the future, leveraging our foothold in gastroenterology and mental health and expanding these same franchises over time. Being solution-centric was also the motive for acquisition of MyPath Melanoma in 2021.
Although we already had our own DiffDx-Melanoma for these difficult to diagnose melanocytic lesions, we felt MyPath could complement our DiffDx-Melanoma test and enable us to provide the most comprehensive molecular testing solutions for these types of lesions and these types of patients. By pairing these two tests, more patients now receive actionable results more of the time.
In fact, over 98% of the time, enabling a more confident diagnosis and clearer treatment pathway for more patients. We've also gone out of our way to lighten the administrative burden placed on our customers. When we first purchased Cernostics in December 2021, they did not have the personnel to secure specimens for testing, slowing the process down and delaying the delivery of clinically actionable test results.
Since then, we have taken the reins of that process and now handle the tissue sample acquisition on behalf of our customers, greatly expediting the turnaround time for actionable test results. Also, securing ADLT status for our TissueCypher test last spring removed a big impediment for our patients as it exempted the test from CMS's fourteen-day rule, reducing disparate healthcare access for our Medicare patients.
Finally, our acquisition of IDgenetix in April expanded our portfolio of testing solutions into mental health and added pharmacogenomics to our testing repertoire.
This acquisition enabled us to offer a testing solution that we believe has the potential to accelerate our impact on patient care in an area of high unmet clinical need, significantly expand our in-market estimated US TAM by approximately $5 billion, and offer incremental value to patients and clinicians over the standard trial and error approach.
With this comprehensive genetic analysis to help clinicians match patients with the right medication and also identify drug-drug interactions to help improve medication response and remission rates, we feel IDgenetix offers a truly compelling and personalized benefit to mental health treatment and is a good example of Castle being both customer and solution-centric.
Now, let's switch gears and remind you of where we believe our focus enables consistent execution. This slide is intended to have you see the patient journey that all patients walk through, from a screening test performed in healthy non-symptomatic patients to tests that assist in diagnostic confirmation, to risk stratification test decisions, to early-stage treatment decisions, and finally, certainly in the case of cancers, of minimal residual disease or recurrence monitoring.
What you should recognize by this slide is that we are laser-focused on providing innovative tests that answer important clinical questions as it relates to supporting correct diagnoses. Improving risk stratification treatment plan decisions and identifying therapies that are more likely or less likely to work in an individual patient. What we don't do is work in, today, the field of screening or late-stage disease monitoring. In summary, these three areas of focus are truly our sweet spot.
This slide is pretty remarkable. It's remarkable that today in mid-2022, our current in-market estimated US TAM approaches $8 billion. It's remarkable because it's worth noting that at the time of our IPO in mid-2019, just three years ago, our in-market estimated US TAM was less than $600 million.
We have, in the ensuing three years, successfully completed a number of development programs that brought us our DecisionDx-SCC test for use in patients with high-risk cutaneous squamous cell carcinoma, as well as our DiffDx-Melanoma test for use in assisting in the diagnosis of suspicious pigmented lesions. Today, our dermatology franchise makes up nearly $2 billion or roughly one quarter of our estimated US TAM.
As you know, we successfully executed 2 acquisitions in the last 12 months with products serving patients diagnosed with Barrett's esophagus as well as mental health conditions. These 2 acquisitions add an additional $6 billion to our in-market US TAM.
We believe that we are now set up to continue to execute in dermatology and gastroenterology and now in mental health, and that our success in the midterm will be a result of profitably growing our recently launched tests in dermatology as well as our acquired assets.
Finally, assuming we are successful in our development program, we expect to launch our innovative pipeline tests aimed at predicting therapy response in patients with moderate to severe psoriasis, atopic dermatitis, and related inflammatory skin conditions.
If successful, this pipeline test has the potential to add an additional approximately $1.9 billion to our current estimated U.S. addressable market. Now let's talk about how we plan to be successful in each of these three franchises. In our first-to-market suite of skin cancer tests, we are primarily targeting dermatologists, followed by Mohs surgeons who are a subset of dermatologists and surgical oncologists.
We also focus on dermatopathologists for our diagnostic gene expression profile test offerings of MyPath Melanoma and DiffDx-Melanoma. Each of these tests provides our customers with clinically actionable data that will effectively inform their treatment plans and make a more meaningful difference in patient care as they work collaboratively with patients to provide the best individualized treatment plan decisions.
Our strategy for driving market penetration in this market also includes evidentiary development or growing the body of evidence supporting the value of our tests. You may recall for DecisionDx-Melanoma alone, we have, since 2015, seen more than 35 peer-reviewed publications support the value of our DecisionDx-Melanoma test, and we expect the more recent data from our NCI-SEER collaboration to be published by year-end.
As I mentioned earlier, one of our guideposts is customer and solution centricity. Our commercial expansion of our sales force as volumes warrant provides the support to providers focusing on education and a greater understanding of our test. Digital marketing to expand awareness and improve customer interactions provides additional support to maximize impact in the market.
In our skin cancer franchise, we have Medicare reimbursement for our DecisionDx-Melanoma test, DecisionDx-SCC, and MyPath Melanoma.
We expect finalization of a recently released draft LCD that would expand coverage to include DiffDx-Melanoma by mid-2023. We are the leader in skin cancer testing. Our first-to-market test for cutaneous melanoma and squamous cell carcinoma were developed in-house. DecisionDx-Melanoma remains today the only gene expression profile test with two indicated uses: risk stratification for predicting the risk of sentinel positivity as well as distant metastasis.
We acquired TissueCypher in December 2021. We will discuss the ongoing integration in just a moment. I want to first review our focus to drive value in 2023 and beyond. These five areas are our flag posts in advancing adoption and bringing this first-to-market innovative test to more patients more quickly. Sales force expansion.
We started following the acquisition in December of 2021 by hiring 14 outside sales representatives to cover 14 territories across the U.S. in January of this year. Based upon 2Q performance, we have recently expanded to 23 territories so that we will be in a strong position of growth going into 2023. Driving awareness through improved messaging.
We knew with the initial quarter or so of being commercially available, there were going to be adjustments to be made to our messaging, to our data displays, to how we interact with our customers.
We've successfully updated our sales tools and marketing materials beginning earlier in the third quarter of this year so that we're entering the third quarter and fourth quarter of 2022 with improved messaging, improved materials that will enable a gastroenterologist to have that aha moment a little earlier in the sales cycle.
We also leveraged the large body of evidence that was developed and published by the Cernostics team, including a recent publication from a Mayo Clinic pooled analysis in the spring of 2022, as well as abstracts that have been accepted at the DDW meeting and ACG meeting going on later this fall. These elements of data development and education will help us drive again the value of the introduction of TissueCypher for use in patients with Barrett's esophagus disease.
We've also improved our GI outreach and education opportunities and expect those to continue into 2023. We hired a dedicated gastroenterology medical director on September 1 of this year, and you'll get a chance to meet him a little later in this program.
We also hired additionally focused individuals to improve the gastroenterology-focused educational efforts and expand our medical affairs medical science liaison team to improve peer-to-peer programs, speaker programs, ad boards, et cetera, all developed to enable acceleration of the introduction, the awareness, the clinical value of TissueCypher test, and to be able to go ahead and have that translate into quality patient care and use clinically.
We also have a mixed focus on both, maximizing leverage where patients are seen today, which is in the community practice environment, as well as making sure that we're touching points in the medically tertiary care centers of the academic centers across the U.S. Both are being targeted. Both are highly responsive to the TissueCypher messaging.
Finally, similar to our efforts in our dermatology franchise, we'll be focusing on investing in digital marketing to drive volume, improve customer engagement, and expand our market segments. Switching gears to mental health, our playbook focuses on these five main areas. One of our key objectives is to strengthen our leadership team and refresh field training.
Having completed the acquisition of IDgenetix in April, we've been focused on integrating staff and getting up to speed, not only in terms of culture, but procedures and operations as well. As we stated before, our goal with IDgenetix is to execute similarly to how we've been successful in dermatology, and we believe that starts with strong leadership, solid support systems, and the proper training to get everyone aligned and working together productively.
As with our acquisition of Cernostics, we believe the integration will take about a year, so that should be completed by mid-2023. Another objective is to focus on the older patient population who benefits, we believe, the most from one of IDgenetix's key value propositions. That is, not only identifying drug-gene interactions, but also drug-drug interactions to help improve medication response or remission rates.
Since the older populations tends to already be on polypharmacy, our drug-drug reporting feature becomes especially critical, and along with factoring lifestyle criteria, truly differentiates IDgenetix when it comes to mental health testing. That's why we think focusing on the older population is the ideal way to leverage the benefits of IDgenetix and drive adoption in a lightly penetrated marketplace.
In the mental health franchise, we will also begin expanding publications from our published randomized controlled trial in order to continue to differentiate IDgenetix from its competitors. Again, the support of clinical studies and trials and their utility in strengthening credibility and driving adoption will be a cornerstone in expanding our market penetration and market share over time.
Furthermore, it will facilitate the efforts of our commercial team by giving them more materials to educate clinicians and build greater awareness of the value in the marketplace. We will target psychiatry practices and other high prescribers of mental health medications and long-term care facilities.
The prospective customer base is made up of these clinicians who are diagnosing patients who are ready to go on therapy, and that is a mixture of some psychiatrists, some high-volume and will include physicians as well as nurse practitioners and PAs.
Another good example is long-term care facilities. Typically, their patients are on polypharmacy, and they are all often suffering from depression. Again, another ideal market for IDgenetix. Finally, similar to our dermatology and GI franchises, we will be focusing and investing in digital marketing to drive volume, improve customer engagement, and expand again into our market segments.
Now that you've seen how we plan to grow our three major franchises, let me talk about the significant progress we've made in preparing and developing our two newest franchises, TissueCypher and then IDgenetix. As part of our strategic growth plans, you may recall that both of these franchises are not expected to be significant contributors of revenue until 2024.
With their very substantial market size, that being roughly $1 billion for gastroenterology and $5 billion for mental health, we certainly expect them to become vital contributors to our business longer term. As such, we've been very active and diligent in preparing to fully actualize their operations and get things underway. I need to reiterate that these acquisitions were made to grow into material revenue contribution in 2024, 2025, and beyond.
They were not made to fill a gap today. As a result, we have the luxury of taking a measured integration approach that we believe will take over a year to complete for each acquisition so that we can get it right without disturbing our focus on our core dermatology business.
Let's talk about TissueCypher, our new test for use in risk stratification in patients diagnosed with non-dysplastic, indefinite, or low-grade dysplasia forms of Barrett's esophagus. On this slide, you can see the strong progress we've made since acquiring TissueCypher in December 2021. In late 1Q 2022, we were granted advanced diagnostic laboratory test status.
This exempts us from the 14-day rule, as I mentioned earlier, and removes what we were seeing as a disparate impact on the ability of Medicare patients to access our tests. In the second quarter of 2022, we began laboratory expansion efforts, including investments in new instruments, staffing, and training ahead instead of the to stay ahead of our expected volume increases. We also delivered 352 TissueCypher test reports in 2Q, compared to 56 reports in Q1.
Substantial growth just quarter-over-quarter, just as we are getting going, by the way. Early in the third quarter of 2022, the American Gastroenterological Association, or the AGA, clinical practice update was released with the best practice advice for utilization of TissueCypher to risk stratify patients with non-dysplastic Barrett's esophagus. We are also executing on planned commercial team investments, expected expansion of the GI sales team, which I mentioned earlier.
We have accomplished all of this prior to reaching our 2023 integration midpoint, which will be in December of 2022. Next year, we expect that our expanding commercial team will be at its optimal productive capacity by the second quarter. As things progress and we execute further on our growth plans, we believe 2024 is when we start to see material contributions coming from the acquisition of Cernostics and their TissueCypher test.
Now, let's move to IDgenetix. We again see strong progress in integrating operations, and we are preparing for future growth. As with our TissueCypher acquisition, we are taking measured integration approaches that will take just over a year to complete and begin to yield fruit. Again, focusing on ensuring that we will receive material revenue contributions in 2024, 2025 and beyond. I'm pleased to say that we are progressing as expected with our plans and have seen some early success.
Success began with Medicare expanding IDgenetix coverage to include additional seven mental health conditions beyond major depressive disorder, effectively expanding our opportunity set to include conditions such as schizophrenia, bipolar disorder, anxiety, ADHD and several others. In the second quarter, we delivered 827 IDgenetix test reports.
Through both 2Q and 3Q, we conducted sales training to align our mental health commercial team with our own proven commercial playbook, while also integrating key internal functions such as marketing, reimbursement and finance. In the third quarter, we placed a national sales director, who was one of our leaders in our dermatology commercial team, to lead the mental health sales team going forward.
Looking ahead, we expect to close the San Diego laboratory and move our IDgenetix operations in total into our Phoenix laboratory by the end of 4Q 2022. Again, these milestones were accomplished before reaching the anticipated midpoint of our integration process, which will be in mid-April of next year. Like with TissueCypher, we expect material contributions on the revenue side from IDgenetix to come in 2024 and beyond.
Until then, we plan to continue with ongoing process improvements and corresponding reductions in COGS.
Hi, I'm Dr. Brent Moody, and I'm a dermatologist and Mohs micrographic surgeon. I practice in Nashville, Tennessee. Professionally, my time is devoted to caring for patients with all forms of skin cancer, and two of the more common conditions that I deal with on a weekly basis are melanoma and squamous cell carcinoma. This is how I got introduced to Castle Biosciences and their gene expression profiling tests, DecisionDx-Melanoma and DecisionDx-SCC.
In this brief video time we have together, I'll tell you how I use GEP, gene expression profiling, as a clinician and how it can be used to help in the care of patients on a daily basis. Let's start off with DecisionDx-Melanoma. A little bit about skin cancer.
Skin cancer is the most prevalent cancer in people, the most common of which are basal cell carcinoma. If you talk to your coworkers, your friends, your family members, people that you know socially, some of them have had basal cell carcinoma. It is that prevalent. 3-4 million per year for basal cell carcinoma. 8 of 10 skin cancers. Of the remainder, squamous cell carcinoma makes up about 20% of skin cancer, so still a large number.
There's a much smaller group of skin cancers that make up less than 1%, this is where we find melanoma. In melanoma in any given year, there's about 100,000 diagnoses of melanoma in patients in the United States. There's other rare skin cancers that we're not going to really talk about at all today.
Of those 100,000 people, that represents about 7,500 melanoma deaths per year. The good news with melanoma is that many patients do well, particularly patients who are diagnosed early and have non-aggressive disease. You see here that the five-year survival is about 93%. That encompasses patients who are diagnosed with melanoma, have initial treatment, which by and large is going to be surgery. Some of those people never need any additional treatment.
Other people will have early on surgery, but then need additional treatment. They may need what we call immunotherapy, they may need radiation therapy or some additional treatment to ensure their survival. This is where gene expression profiling in the melanoma space can help us, can help me as a clinician who treats these conditions.
It can help me identify those patients who are most at risk for having their melanoma spread, what we call metastasis. Those patients who may need to have immunotherapy or systemic therapy for their melanoma, something beyond just the surgical procedure. The DecisionDx product gives me two pieces of information that help me manage these patients. The first, it gives me the likelihood that a patient is going to have cancer in their lymph node.
There's a very common test that's used for melanoma patients called sentinel lymph node biopsy. It's a test where we can inject a radiotracer dye at the site of the cancer. That dye will then migrate to the draining lymph nodes. Using some equipment, we can determine which lymph node is most likely to have cancer in it if cancer is present.
This is called a sentinel lymph node biopsy test. This has been routinely done for well over 40 years. It has a lot of advantages. It can find small amounts of melanoma that things like CT scans or even PET scans, which is a more advanced imaging technique, cannot detect. There are instances where the cancer is in the lymph node in such a small amount that it cannot be detected except microscopically. We biopsy that lymph node, and a pathologist examines it for us.
Traditionally, we have decided who receives this test based on 2 pieces of information, the depth of the melanoma, that is how deep it's going, when it's measured, and whether or not it exhibits a feature called ulceration, which is, again, something the pathologist will tell us. For years, that is how we decided.
The bad news with the sentinel lymph node biopsy test is that the vast majority of times they're negative. About 85 times out of 100, when a patient has the test, there's no cancer detected. The question is, can we better refine and do a better job of who we select in undergoing the sentinel lymph node test? It's generally a safe test. Many of my patients have it on a monthly, sometimes weekly basis, and the vast majority do well. There are side effects to the test.
There can be infection, there can be nerve damage, there can be swelling, and there is cost associated with this test. One of the things that DecisionDx-Melanoma does for me is help me figure out which of my patients are more or less likely to benefit from having the sentinel lymph node test.
The way this was developed was using the genetic profile of the melanoma, plus other features that we know are important in predicting the likelihood someone's gonna have a lymph node be positive. That's the thing I mentioned earlier, the thickness, the ulceration, what we call the mitotic rate, or how active the cancer is, and the patient's age. This gives me additional information that can help me better select which patients need sentinel lymph node biopsy.
This may identify patients who traditionally we would not have thought needed sentinel lymph node biopsy, but the genetics proves that they're a high-risk patient who would benefit from the test. Or it may find patients who traditionally we may have thought would benefit from the test, but in fact, are low risk enough that we can avoid this test altogether.
The second piece of information that this test gives me is what is the likelihood that this particular patient's tumor is going to metastasize, that is spread to other places in the body? What's the likelihood that it's gonna recur, and what's the likelihood that it's going to kill the patient or have mortality? That is the information that DecisionDx-Melanoma can give to clinicians who treat melanoma patients.
The big question is, well, why does this test need to exist? Clearly, prior to this test being available, we had ways to predict which patients were more or less likely to do well. I wanna take a couple of minutes to teach you a little bit about the history of melanoma staging and why this test now exists. First, let's talk a little bit about staging.
Staging is our way of putting cancers into categories, putting it into risk categories. Is this a low-risk cancer? Is this a high-risk cancer? Is it somewhere in between? Why do we do that? Why do we stage a cancer? It's to do several things. It does help predict the biologic behavior of that cancer. These are things that patients wanna know.
When I have a patient that I diagnose with melanoma, they wanna know, what is the likelihood that this melanoma is going to kill me? What's the likelihood that it's going to spread, and I'm gonna have to have a treatment with a medical oncologist? Or what's the likelihood that this cancer is gonna come back? It helps us predict that behavior. It also helps us determine the appropriate treatment. Based on a patient's stage, we may offer them systemic therapy.
We may offer them to have imaging. It determines the extent of the surgery they have. Staging helps determine those things as well. The ultimate reason why we stage patients is we wanna alter that outcome. We don't only want to be able to predict it, we want to be able to understand which patients are high risk, and are there things that we can do for those high-risk patients that decrease the likelihood of them having a poor outcome.
This is why staging and accurate staging is critically important in cancer treatment. Thinking about melanoma, there are some problems I see with the current staging. The current staging. A lot of people who are Stage One. Stage One are thin melanomas, patients who do not have it in their lymph node. Stage One people, by and large, do well.
80% of the melanomas diagnosed in the United States every year are Stage One. That's the good news. We've done a good job of finding these melanomas early, and we know the earlier we find them, the better off patients are. There are a lot of deaths that come out of this Stage One group. Of that 100,000 people that get diagnosed with melanoma each year, about 80%, as I said, so 80,000 are gonna be Stage One.
Of the people who die of melanoma, about one in four are Stage One at the time of diagnosis. A lot of people who are Stage One, who otherwise we think should do well end up dying. Well, what's the explanation for that? Why is it that so many Stage One people do well, but some do poorly?
What can we do to help figure out which of these Stage One patients are more likely to do poorly than others? Another big problem we see is that Stage Two B and Two C patients, and the details of what that means are probably not critical here, but they do worse than Stage Three patients. This is a big problem. Stage Three patients already have cancer in their lymph node, where Stage Two and Three patients do not. We have instances where Stage Two patients do worse than Stage Three.
In cancer staging, it should not be that way. Stage One patients should do better than Stage Two patients should do better than Stage Three patients, who should do better than Stage Four patients. That's in the ideal world. It really raises the question of can we do better?
Can we update our thinking in melanoma staging? The history of melanoma is very interesting. We could have a whole lecture on that sometime. Even folks who are not in medicine find it interesting. For the vast majority of the time that we have known about melanoma as a distinct entity, it was a mystery. It was really hard for physicians and surgeons to predict who was going to do well, who was going to do poorly.
As a result, for many years, really up until the 1980s, the surgical treatment of melanoma was quite extensive. Very wide, very deep, prophylactic removal of many lymph nodes. It's really unclear whether any of this very intense treatment, which had with it a lot of downsides, was how effective it was.
It really begs the question of where are we with this? Can we update our thinking? The first attempt to understand melanoma better occurred when Dr. Wallace Clark, who was a pathologist, introduced this concept called Clark's Level. It was a way to look at how deep the melanoma was based on anatomy.
You know, at what level of the skin did that melanoma extend to? Love this quote from Dr. Clark's paper. "It has been quite difficult to explain this apparent striking discrepancy in the survival rates of malignant melanoma." People were struggling to explain why did some people do well and some do poorly in this disease. Wallace Clark's work did really push things forward in helping understand and predict which patients are going to do well and which are gonna do poorly.
Even at this stage, though, I will point out one thing to you. If you look at the level here. Patients who survived had lower level than patients who had advanced level. If you were level five, you're much more likely to die than if you're level two in looking at it the other way.
What this work could not do is explain, well, why do most people who are Stage Two do well and some do poorly? It did a good job of explaining the likelihood that someone was going to have a problem based on the level of invasion, but could not explain what we call intra-stage variability. Had no way to explain, okay, why do some of these Stage Five people do well? Why do some Stage Five people live where most die? No explanation for that.
Another advance came one year later in 1970. Dr. Alexander Breslow introduced what's called Breslow Thickness. If you recall from a few slides back, you saw those words on that slide. Dr. Breslow simply measured, not like Dr. Clark did based on, you know, what level of histology it went to, but just the measurement in millimeters. How deep was it going? He saw that the deeper it was in millimeters, the more likely the patient was to have a recurrence or a metastasis.
Again, he could explain the difference between thin and thick in outcome, but really couldn't explain, well, why do some of these people with a greater than 3-millimeter melanoma have a recurrence or metastasis and some don't? No explanation, again, for what we call intra-stage variability. We're explaining why patients of the same stage have different outcomes. Then Dr.
Breslow opinion for cutaneous melanoma is a most unpredictable lesion. Finally, our history lesson is coming to a close, Dr. Charles Balch at MD Anderson realized that this phenomenon known as ulceration, which is a pathology finding, was again correlated with outcome. Patients who had ulceration did worse than patients who didn't have ulceration. Which brings us to our current staging, the current AJCC, American Joint Committee on Cancer, and that AJCC is the language of cancer.
It is what medical oncologists, radiation oncologists, surgical oncologists use to express a cancer stage. This is how it's made up now. The Breslow thickness we talked about from Dr. Breslow's work in 1970. Ulceration in 1980 was described. It was added to the staging in 2001.
You know, the Breslow was not really added in a meaningful way until 1983. There's a lag in these groundbreaking discoveries and when they get added into the official staging system. The final things to round this out is, well, what is the lymph node status? That's where that sentinel lymph node biopsy comes into play. Then distant metastasis. Has the melanoma spread to the liver, the lung, the brain, things of that nature.
There are other things that used to be used, including Clark's Level, but no longer that influenced staging at times in the past. Here's where we currently are, AJCC 8. We think about current melanoma staging, and it relies upon factors that are getting to be pretty old. Ulceration was described 40+ years ago and Breslow 50+ years ago. These are still critically important.
This really speaks to the amazing work that those physicians did. But again, back to our key questions. What explains the variation in outcome within a stage? That intra-stage variability? Again, why do some people with stage one die when most do well? Why do some people with stage three cancer in their lymph node do well and many die? So that's one big question that is unanswered with traditional staging.
Again, I've said this before. You know, why do some people with bad cancer seem to beat the odds and some people with good cancers or cancers we think should do well do poorly? In my view, and this is where the genetic profiling in melanoma really helps me as a clinician figure out using the traditional risk factors plus this new piece of information.
When new, I say new, it's been out for several years now, but this additional information can really help me figure out, you know, within a stage is my patient more or less likely to do poorly? It helps me figure out which of the patients perhaps need more intensive therapy, more intensive monitoring, to help them do well. This gene expression profiling better helps me predict the biologic behavior of the tumor. In other cancer types, this gene expression profiling is standard practice.
Uveal melanoma, breast cancer. It's gaining traction in prostate cancer. I think we're gonna see more of this in the future. In the past, we relied on histology or what the pathologist told us. Now we have histology plus the biology of these cancers to help us figure that out. I've alluded to some of this already.
You know, traditional approaches to staging might miss patients with aggressive behavior of their tumor. Those patients who are early stage at the time of diagnosis, again, most of them are going to do well. How can we identify those patients who are not going to do well? If you look at this slide here, you'll see a patient with what we call a stage 1B tumor that's still considered a low-risk tumor. It's relatively early. Many of those patients are gonna be cancer-free.
Some of them will progress on to what we call stage 4 or metastatic disease. If we can identify patients who are more likely to go on to have metastasis and find those metastasis early, patients do well. Immunotherapy, which is an amazing advancement in melanoma.
If you watch TV, you've seen ads for some of these drugs. They are more effective in early metastasis. By the time someone's metastasis has really spread through their entire body, even the amazing immunotherapy we have today is less likely to be effective. What can we do to make sure we're identifying these high-risk patients? In my opinion, this genetic profiling is one of the things that we can do. These are called Kaplan-Meier curves, and they reflect some sort of outcome.
I'll explain to you briefly so you understand it. You can see this is recurrence-free survival or distant metastasis-free survival. What is the likelihood that someone with a certain feature to their melanoma is going to not have a recurrence? They're gonna be free from recurrence. Recurrence-free survival.
If you look at both of these, that top blue line are patients with favorable genetics. Patients with what we call a Class 1-A result. That bottom orange line are patients with unfavorable genetics. The worst possible genetics, what we call a Class 2-B result. You can see in both instances, the proportion of people who are recurrence-free or free from metastasis is much lower in patients with bad genetics.
If you have bad genetics, you're much more likely to have a recurrence, much more likely to have a metastasis than if you have favorable genetics. The dashed blue line is what we call a 1-B result. That's less favorable than 1-A, but not as bad as a class 2 result. That orange dashed line is a class 2-A result.
You've got blue lines are class one, orange lines are class two. One being the good result, two being the bad result. Thinking about melanoma and the things I've talked to you about, we wanna find patients who are at high risk, so we know we can watch those high-risk patients more closely, identify their recurrence or their metastasis as early as possible to give them the greatest likelihood of surviving. That's the premise behind what I've told you so far.
The GEP, in my opinion, does that. I use it to help risk stratify patients. Those Kaplan-Meier curves I just showed you clearly demonstrate that the genetics are prognostic. They predict the outcome. Nothing I've shown you so far actually demonstrates that testing a patient with the DecisionDx-Melanoma makes them live longer.
It goes back to what I earlier spoke about, that ideally we want to alter patients' outcomes. What Castle Biosciences has done is they partnered with the National Cancer Institute and their epidemiology database. It's called the SEER Registry, Surveillance, Endpoints, Epidemiology Registry.
The NCI just looks at cancer patients and follows them and figures out, well, how many of them are alive, how many of them are dead? What are some features that each of those groups had? Castle Biosciences partnered with the National Cancer Institute to try to answer the question, does testing with GEP DecisionDx-Melanoma improve the likelihood that a patient's going to live? What the result showed is that indeed it does.
Testing melanoma patients with DecisionDx-Melanoma showed what we call a survival advantage.
This top pane is what we call overall survival, and that's just is the person dead or not dead? Are they dead or alive? Then melanoma-specific survival boils it down to, okay, did the patient die of melanoma, yes or no? In each instance, you can see a decrease in the risk of death in patients who were tested with DecisionDx-Melanoma. Overall survival at 21% decrease in death risk. In melanoma-specific survival, a 27% decrease in death risk.
I don't wanna get too deep into statistics, but any of you who studied statistics, or do this for a living, you'll notice on the right-hand side it talks about a P-value. A P-value is a measure of is the difference that we're observing statistically significant? Is it a meaningful difference?
The threshold we look for in medicine is any P-value less than 0.05 is considered to be clinically meaningful, so it's statistically significant. In both instances, these were proven to be statistically significant. Really for the first time, we can now say that testing a patient with DecisionDx-Melanoma decreases the likelihood that they will die, and specifically that they will die of melanoma. This is very helpful information for those of us who treat this disease clinically.
Additionally, you know, patients who have been tested and have been queried about this, they have been happy that they were tested. The patients don't regret being tested. They want to know this information. They wanna know as much as they can about the likelihood that they're going to do well or they're going to do poorly.
Again, just as I find it useful as their physician to have this information, patients today want to be informed. In my view, they wanna be active participants in the decision-making of their care, and this is another tool that I think can empower them to make decisions that are in their best interest based on their goals and desires, for their cancer treatment. That's wrapping up melanoma. Now I wanna transition to DecisionDx-SCC.
Many of the concepts or ideas I spoke about in DecisionDx-Melanoma really apply to DecisionDx-SCC. You know, the same idea, we want to predict, well, what is the likelihood that this patient's going to do poorly versus not do poorly? The chief outcome that we look at with DecisionDx-SCC is metastasis.
What is the likelihood that this particular squamous cell carcinoma is going to metastasize? That means it's going to go to a lymph node or it's gonna go somewhere else. When we look at squamous cell carcinoma, it's a serious problem. It's a very prevalent disease.
Remember I told you that 20% of skin cancers are squamous cell carcinomas, and depending on what database you look at, there's anywhere from 3, 4, maybe even 4.5 million skin cancers per year in the United States. There's a lot of squamous cell carcinoma. There's a lot of deaths. This says 15. It should say 15,000.
There's about 15,000 deaths from squamous cell carcinoma, about 7,200. Here it says 7,200. Earlier, I said about 7,500. The point is, more people die of squamous cell carcinoma than melanoma.
This is an important disease. Just like for melanoma, we have staging systems that are designed to help us make decisions based on the patient's risk of having a bad outcome. The staging really, it's where you want to consider high risk. There are multiple ways of looking at squamous cell carcinoma. If you have a very strict criteria for high risk, your high-risk bucket of patients is going to be smaller, but you may leave people out.
There may be people who indeed are likely to metastasize that we don't recognize as high risk because we're making our criteria too strict. If we make our criteria over here very loose and consider a lot of people high risk, then there's going to be a lot of those high-risk patients who never have a problem.
We really have to figure out, well, where do we draw the line for high risk? Again, this is something where GEP can help me in my practice. Let's look at our current SCC staging options. Just like I talked about with melanoma. Well, why does this test exist? Why did someone bother to invent this test? We have current staging systems that are designed to predict outcomes, and there are really. Well, I'm being a little facetious here.
There are three, and I put a fourth in there, gut instinct. Those of us who treat this disease commonly, we just develop a sense that, okay, this one is not good. Our gut will tell us, something's not right. There's the AJCC. That's what we talked about earlier with melanoma. And again, it is also used for squamous cell carcinoma.
There's a second one that's been developed called the Brigham and Women's Hospital system, and a third one, the National Comprehensive Cancer Network. There are three official staging systems, and then gut instinct. The first thing I'll tell you, whenever there are multiple options for something, it's because one is not clearly better than the other. You know, if one of these was just so much better than the others, the others wouldn't exist.
Let's take a quick look at these staging systems, and I think it will help you understand why clinicians like myself utilize the GEP test in squamous cell carcinoma, DecisionDx-SCC. Let's dig a little bit deeper into these staging systems. This is not something you need to know.
It's just this is the AJCC staging system, and it involves really like how big is the tumor, how deep is it going? Similar things we talked about in melanoma. Is the cancer around a nerve, which we know is bad. It takes all of these factors, and comes up with a staging system.
The question is, well, how well does this help someone like me make decisions? Well, let's look and see how well AJCC performs in predicting an outcome. In this case, the likelihood that a particular cancer is going to metastasize to a patient's lymph nodes. Again, this is similar to that idea we talked about earlier, these Kaplan-Meier curves. What this is showing is that a patient with a T3 squamous cell carcinoma. Don't worry about what that means.
It's a bad one, but we don't have to worry about the details. Over the course of three years, has it about a 13, 14, 15% chance of having a metastasis. Then it shows this T2 patient, which is a completely different stage, has like a 12%-13% chance. The first problem with AJCC is here. These two, there's a distinction. We make a distinction between T2 and T3, but there's no difference in how well those patients do. One problem with AJCC.
The bigger problem for me as a clinician is this big red arrow here, and this tells me that, again, the vast majority of my patients who are AJCC T3 never have a problem. If I'm going to say, "Okay, you have a T3 squamous cell carcinoma, which is a bad cancer, I'm going to do something to you.
I'm going to send you for radiation treatment, or I'm gonna send you for CT scans or whatever." 85 times out of 100, the patient didn't need that. Again, what's my problem is I don't know which patients might benefit from that. I'll also point out this little T1 group that we generally think do well. Again, not all of them do well, and some of them will have a metastasis. Similar to melanoma, we have a lot of stage 1s. A fair number of metastasis will come out of this stage 1 group.
Even though the likelihood is low, the fact that there are so many of them means that about 30% of people who ultimately metastasize were early-stage at the time of diagnosis. Well, maybe Brigham and Women's, the other system will help us. Not really.
It does better here between these two, T2A and T2B. We get a distinction here and a difference. Again, the vast majority of patients, even with bad cancer, T2B, never have a problem. I'm stuck as a clinician. How do I figure out what I'm gonna do with these patients? Lastly, the NCCN has a guideline, a little more comprehensive. It's newer, so I don't have good performance data on it yet. Maybe this is gonna be the savior. I don't know. Staging challenges. I've already mentioned these.
What are my solutions? I could just decide, okay, I'm just gonna observe everyone. I'm gonna treat you with surgery, which is the main treatment for this disease, and just observe you and wait and see what happens. I'm gonna under-treat a few people to avoid over-treating a bunch. I could treat everyone.
I could say, "Look, if you have a bad cancer, I'm gonna do all these things to you," knowing that I'm over-treating a lot of people to avoid under-treating a few. I think you're starting to see my problem as a clinician. Well, what's the solution? Let's refine our prognostic ability.
What other tools are available that will help me figure out, of these high-risk patients, which are the ones that are actually most likely to have metastasis and warrant additional treatment, additional surveillance, imaging, things of that nature? That's where DecisionDx-SCC comes into play. Similar technology to the melanoma test, different number of genes, and slightly different results. Instead of having the four categories of results, we only have three, but that's okay.
It helps me figure out which patients, again, are going to be more likely to have a problem. In fact, there's evidence to suggest that the genetics may be the most predictive feature in helping decide if a patient's more likely to have metastasis. More predictive than some of the traditional risk factors that are used by AJCC, Brigham and Women's Hospital, and the NCCN. You can see Class 1 patients have a relatively low risk of metastasis. Class 2A patients, about a 20% risk.
That's getting to be significant. Class 2B patients, a very high risk of metastasis. Again, this helps me figure out which of these patients are more or less likely to need those next level interventions. It's been looked at and studied. Again, figuring out what we call risk-aligned management.
When this has been presented to my peers, we are recognizing that we should treat patients differently based on the genetics. This is, as the slide said, what I call actionable data. It is a piece of information that I can use to help design customized treatment programs for a patient, again, based on the traditional risk factors, but also now based on the genetics.
When we call what we call real-world use data, you know, looking at how this happened after the product was launched and people like me were ordering the test. You know, many patients that were ordered had more than one risk factor. My colleagues recognized that this is for high-risk patients, and most of the patients were either NCCN high risk or very high risk.
The good news is my clinician colleagues appear to be using this test as it was designed, for high-risk melanoma patients. When presented with scenarios and asked, "Well, what are you going to do differently based on the DecisionDx-SCC data that you get?" what we see is this is just one example of many that are available, that if a patient has favorable genetics, so less likely to have a metastasis, my colleagues said, "Hey, we're maybe gonna see that patient less frequently in the office."
For patients, that can be a big deal. The difference between being seen every three months and every six months for them is an important thing, less disruptive to their life, less money out of pocket for them, less inconvenience.
My colleagues were less likely to send patients for CT scans or ultrasounds or MRIs, some sort of imaging study, if they had favorable genetics. Less likely to recommend radiation after surgery if they had favorable genetics. Whereas if they had unfavorable genetics, they were more likely to do each of those things. My colleagues are intuitively understanding how to incorporate this data into their decision making.
Similar to melanoma, where a patient has unfavorable genetics, we're going to do more for that patient. Squamous cell carcinoma, unfavorable genetics, we're going to do more for that patient. If they have favorable genetics, we can de-escalate or dial back their care.
I hope this little really brief introduction to some of the science of GEP and how a clinician such as myself views this, how I use it, helps you understand the importance of this test for patient outcomes. Again, my name is Dr. Brent Moody, and thank you for your patience. I really appreciate the opportunity, albeit remotely and virtually, to share with you some of my thoughts about Castle Biosciences' DecisionDx-Melanoma and DecisionDx-SCC. Thank you.
My name is Matthew Goldberg. I'm a board-certified dermatologist and dermatopathologist and medical director at Castle Biosciences. Today, I'll be speaking about the diagnostic gene expression profile tests, DecisionDx- DiffDx-Melanoma, and MyPath Melanoma. Two diagnostic gene expression profile tests offered by Castle Biosciences, as well as discussing the inflammatory skin disease pipeline program ongoing here at Castle.
To start the discussion about diagnostic gene expression profile testing, and specifically the MyPath Melanoma and DecisionDx- DiffDx-Melanoma tests, I think it's important to outline the scope of the clinical problem, specifically the unmet clinical need in patients who have difficult to diagnose melanocytic skin lesions.
These lesions represent a significant clinical challenge for dermatologists and dermatopathologists, as many patients who present to the dermatology clinic for evaluation of concerning skin lesions require a biopsy, a diagnostic skin biopsy, to help establish whether they're faced with a benign melanocytic nevus or a malignant melanoma.
This level of patient concern leads to many visits to dermatologists and other skin cancer providers across the country, leading to almost 2 million biopsies of suspicious pigmented lesions across the United States.
While most of these skin lesions can be accurately diagnosed as benign nevi or malignant melanoma using the tools that dermatopathologists have on-hand to differentiate between those two entities, there are approximately 300,000 skin lesions that can't be confidently classified as either benign melanocytic nevi or malignant melanoma using the traditional methods of histopathology.
These difficult to diagnose melanocytic skin lesions have been well documented as a challenge for dermatopathologists and dermatologists in the literature. Despite access to subjective second opinions and other diagnostic ancillary tests, this issue of diagnostic uncertainty remains a significant challenge for a portion of lesions with a high rate of discordance that's been noted between subjective reviewers of difficult to diagnose cases as seen here.
This diagnostic ambiguity is impactful and can lead to both diagnostic uncertainty and clinical management uncertainty that has significant consequences for patients who have these lesions.
Specifically, when the diagnosis of whether a lesion is benign or malignant is uncertain, it can lead to significant over-treatment, which has increased patient morbidity and cost to the system where benign skin lesions are actually unnecessarily excised when they could be confidently not re-excised if the diagnosis was that of a benign lesion. On the other side, it can lead to significant under-treatment.
A lesion that would have been best characterized as melanoma is not treated as a malignant melanoma, which leads to the potential for missing the diagnosis of a malignant melanoma. Essentially, accurate diagnosis of benign or malignant neoplasms for these challenging melanocytic processes is important for providing the right diagnosis and then the right treatment plan to patients with these challenging lesions.
MyPath Melanoma and DiffDx-Melanoma are both diagnostic gene expression profile tests that are validated to classify these ambiguous melanocytic lesions as suggestive of benign or suggestive of malignant neoplasms. The current laboratory workflow at Castle Biosciences leverages the strengths of these two gene expression profile tests to help guide better patient care.
MyPath Melanoma can be followed by DiffDx-Melanoma if an intermediate or multiple gene failure result is rendered in the current workflow. This utilizes the strength of the clinical evidence supporting the clinical validity and clinical utility of both assays as both tests are in-house at Castle Biosciences. First describing background of the MyPath Melanoma test. This is a gene expression profile test that evaluates the expression of 23 genes, including two variants of the PRAME gene.
There have been over 35,000 lesions tested in a clinical setting with over 1,300 patients in clinical validation studies to date. On DiffDx-Melanoma, this test uses a neural network algorithm to evaluate the expression of 35 genes and has been validated in a wide variety of melanocytic lesion subtypes and has a very low rate of intermediate test results.
Diagnostic gene expression profile testing here at Castle is designed to provide clinically actionable and objective results for nearly all patients whose lesions are submitted for testing. MyPath Melanoma and DiffDx-Melanoma, when positioned in our current workflow, lead to over 98% of patients receiving a result of either suggestive of benign melanocytic neoplasm or malignant melanocytic neoplasm.
Which means that the overwhelming majority of patients who have their lesions submitted for testing will receive this objective result that it can be combined with the clinical, histopathological, and other laboratory information to provide the most value to lead clinicians to a definitive diagnosis to allow for better patient care for those individuals who have been diagnosed with these challenging melanocytic neoplasms.
Diagnostic gene expression profiling has demonstrated clinical utility for both dermatopathologists and dermatologists, and there have been multiple peer-reviewed publications that show the utility for both of these specialists for the benefit of patient care.
When used by dermatopathologists, diagnostic gene expression profile testing can reduce the rate of diagnostic ambiguity for pathologists interpreting skin biopsies and can reduce this rate of uncertainty in lesions that are currently signed out with significant diagnostic uncertainty.
On the clinical side, dermatologists who receive results with diagnostic gene expression profile testing have reduced re-excision recommended for patients with benign gene expression profile results, which results in fewer unnecessary re-excisions.
In a study where patients have been managed using diagnostic gene expression profile test results and patients who receive a benign result subsequently forego surgical intervention, no adverse events were seen in the study referenced here on the right. Essentially, diagnostic gene expression profile testing led to a significant reduction.
Over 80% of patients did not have lesions with a benign gene expression profile test undergo re-excision without any adverse events in the follow-up for this study period. The next two slides contain a case study that highlights the clinical application of diagnostic gene expression profiling for a patient with a difficult to diagnose melanocytic neoplasm.
The case relates to a 27-year-old woman who was initially seen by a primary care physician for a concerning lesion on her back and was referred to an expert dermatologist for evaluation of this pigmented lesion. The patient of note had a family history of melanoma and pancreatic cancer, and so the lesion was of heightened concern.
On clinical examination with the dermatologist, she was noted to have an asymmetrically pigmented 6 mm lesion on her back with atypical features on both dermoscopy and confocal microscopy.
These atypical findings led to the differential diagnosis before the biopsy to include benign diagnoses such as an atypical nevus with regression, as well as malignant diagnoses that included melanoma in situ or malignant melanoma. A diagnostic biopsy was pursued in this case and sent to the pathologist.
Upon histopathology review, there were features that were concerning for malignancy, such as focal confluence of melanocytic nests, as well as features of a benign nevus, leading to the final diagnosis of a compound nevus with atypical features. There was a recommendation as well for close clinical follow-up on the initial pathology report.
However, due to the conflicting histopathology and clinical features, with the recommendation for close follow-up, it was unclear for the dermatologist how to best proceed for this patient.
Is this lesion best characterized as a compound nevus with atypical features, or could this actually represent more concerning melanocytic process such as a malignant melanoma? After a conversation between the dermatologist and the dermatopathologist, the MyPath Melanoma diagnostic GEP test was ordered, which returned a result suggestive of a malignant neoplasm.
This result, when incorporated with the other clinical and histopathology information, led to a revision in the pathology report, an addendum report, where the diagnosis of melanoma in situ could not be excluded. This modification led to a revision in the patient's treatment plan, a re-excision of this spot, as well as an increase in follow-up now four times per year.
At this heightened follow-up of quarterly visits to the dermatologist, in that follow-up, two additional, evolving pigmented lesions of concern have been identified in monitoring.
I think that this case highlights how the addition of gene expression profiling in the setting of other tools that dermatologists and dermatopathologists have access to today can add value to patient care, move from diagnostic uncertainty and ambiguity towards clearer diagnoses with clear management plans for patients that can identify concerning lesions, provide definitive treatments for them, and route patients to appropriate monitoring for their impactful skin disease.
Moving on now to discuss the inflammatory skin disease pipeline test, specifically the pipeline test to predict response to systemic therapies for patients with moderate to severe psoriasis, atopic dermatitis, and related conditions with a target launch date by 2025. Much of the discussion today and my prior comments have been based on areas of dermato-oncology, where patients are faced with malignant processes on their skin.
However, there are also impactful inflammatory skin diseases that dermatologists and their patients face, every day in the clinic. There's significant unmet need in patients with inflammatory skin diseases that have a high burden to patients and a high cost to the healthcare system.
Most commonly identified among those inflammatory skin diseases are psoriasis and atopic dermatitis, which are among some of the most frequently seen skin rashes in dermatology clinics across the country. These diseases are impactful for patients who have these disease st ates, and the treatments are significantly different for both psoriasis and atopic dermatitis.
For patients with moderate to severe disease can include costly medications that have high year-to-year costs, but are also used for the lifetime of those patients who are affected by these skin conditions.
Importantly, there are some clinical mimickers and cutaneous T-cell lymphoma, CTCL, can mimic the presentation of psoriasis and atopic dermatitis. It's important to highlight that systemic therapies for atopic dermatitis and psoriasis are currently prescribed using a trial and error type approach, which I'll explain here on the next slide.
What's important to highlight here is that systemic therapy guidance tools or precision medicine tools have the potential to streamline therapeutic interventions for patients to avoid ineffective and expensive medication courses that include multiple switches in medications prescribed for these patients.
On this slide is a schematic of a patient journey for an individual who's affected by one of these, inflammatory skin diseases. First, a patient here on the far left would present to the dermatologist with the signs and symptoms of an impactful skin rash.
These rashes can involve large surface areas of the body in impactful areas such as the palms and soles and face. A clinical diagnosis is made based on the signs and symptoms that these patients present with, and doctors can accurately diagnose patients with having psoriasis, atopic dermatitis, or a related condition.
The treatment often begins with focused topical therapies or light therapy for patients with these skin diseases. However, based on the severity of their skin disease, the decision can be made to pursue systemic therapeutic agents for either of these conditions. The management plan at present is determined on the clinical features that are present, and then this trial-and-error loop seen here on the right based on the patient's response to therapy.
The specific therapies are guided based on the clinical diagnosis, but the likelihood of therapeutic success does not impact the decision to pursue a specific medicine for a specific patient, but rather a trial-and-error approach is used for these patients, which includes the starting and stopping of multiple new medications for those initiated on systemic therapy.
As a result of this current approach and unmet need to identify a likelihood of systemic therapy response, Castle has started two studies to aid in the treatment of inflammatory skin diseases. First, the IDENTITY study, which is to help guide therapy selection for atopic dermatitis and psoriasis, which is a prospective re-enrolling multicenter study using a non-invasive skin scraping method to obtain samples as part of the study.
I'm also introducing today the SIGNAL-MF study, which is a study to identify mycosis fungoides, which is a type of cutaneous T-cell lymphoma that can mimic atopic dermatitis and psoriasis. Again, these samples will be obtained through a non-invasive collection method, and it's prospectively enrolling multicenter study with 15 sites targeted, with 13 sites that have committed to enrolling patients to date. First on the IDENTITY study.
This is Castle's inflammatory skin disease pipeline test. It's being developed to predict a patient's response to systemic therapy when started on systemic therapy for atopic dermatitis, or psoriasis. I'd like to highlight that 55 sites across the country have committed to enroll in this study, with over 330 patients enrolled to date.
We've hit some program milestones already with a steering committee composed of national experts in the treatment of psoriasis and atopic dermatitis in 2021, with first patient enrolled at the end of 2021, and we had a proof of concept that was completed earlier this year that I'll explain in the next slide. I'd also like to point out that the product launch date is targeted for 2024.
The proof of concept that's been presented at a national dermatology conference relates to the ability of a non-biopsy tissue collection method, in this case scraping, for this inflammatory skin disease pipeline to be able to identify differential gene expression from this non-invasive method.
Seen here in these photos is a demonstration of a simple-to-use, non-invasive skin scraping method using tools that are available in every dermatologist's office with a technique that is easy, easily adaptable by any dermatologist to be sent to Castle Biosciences in an RNA-preserving buffer. When done in this way, the laboratory at Castle Biosciences has been able to demonstrate gene expression from these non-invasive skin samples, seen here on Figure A.
In Figures B and C on the bottom, from the RNA content submitted to the laboratory, differential gene expression from affected areas and non-affected areas of atopic dermatitis and psoriasis has been demonstrated. This proof of concept milestone is significant, demonstrating that this non-invasive method can lead to differential gene expression for the development of a test for this unmet clinical need.
The SIGNAL-MF study, Castle's inflammatory skin disease pipeline test, could include a diagnostic ancillary test to identify mycosis fungoides. This study has targeted 15 sites with a study size of 500 patients. Earlier this year, the protocol was finalized, the IRB was approved, an investigator meeting was held, and again, this is part of the 2025 targeted launch date for this inflammatory skin disease test.
On this last slide, I'd like to present the revised patient journey that maps the experience of patients moving through the dermatology clinic after the clinical availability of a therapy guidance test from Castle Biosciences for patients with psoriasis and atopic dermatitis.
Here on the far left, when a patient presents to the dermatology clinic and is diagnosed with moderate to severe psoriasis, atopic dermatitis or related condition, and the decision to pursue a systemic therapy is made, instead of immediately starting on the trial-and-error treatment loop as is done today.
The clinician can order Castle's innovative test to support systemic therapy selection decision-making using tools that are readily available in the clinic, Castle Biosciences test under development, to have management plans be determined based on an individual's biologic profile, using this personalized medicine tool.
The goal here is that when a patient is started on systemic therapy for psoriasis and atopic dermatitis, that they'll start with a higher likelihood of treatment response, effectively avoiding the trial-and-error treatment loop, resulting in improved patient outcomes and quality of life.
At Castle, we believe that personalized therapy guidance can lead to reduced medication switches for patients with psoriasis and atopic dermatitis and related conditions and significant healthcare savings as a result. Our test is expected to be launched in phases, starting first with systemic therapies most commonly prescribed in the IDENTITY study cohort. With that, I thank you for your attention.
Esophageal adenocarcinoma is an aggressive cancer associated with poor outcomes. The only known precursor condition to its development is Barrett's esophagus. My name is Craig Munroe, and I am the Medical Director for GI at Castle Biosciences.
I am a gastroenterologist with over 10 years of experience in the field. Thank you for the opportunity to present on our first product in GI for Castle, TissueCypher. TissueCypher is the first and only precision medicine test developed to predict the development of esophageal cancer in patients with Barrett's esophagus.
The test combines spatial biology, digital pathology, and validated computational analyses to create an assay which is being supported by numerous peer-reviewed publications. The end result is the TissueCypher Barrett's esophagus assay. The vast majority of the 435,000 annual endoscopies done with a finding of Barrett's esophagus are found to have non-dysplastic disease.
The clinical challenge we face is that most patients will not develop cancer in their lifetimes. Our dilemma is to predict who will progress in order to tailor our care for our individual patients and avoid harm.
While progression rates for this condition are generally low, 50% of annual progressors are initially diagnosed with non-dysplastic disease. Given the limitations of pathology review, patients diagnosed with low-grade dysplasia are down-staged up to 85% of the time when the specimens are re-reviewed by expert GI pathologists.
Lastly, and perhaps even more troubling, is that 25% of high-grade dysplasia patients or cancer patients' diagnoses occur within 1 year of a previous endoscopy, indicating that the patient was harboring prevalent disease at the time of their procedure. Prevalent disease is defined as progression to high-grade dysplasia or cancer within 12 months of endoscopy. In short, our current standard, which is histology, is limited in its ability to identify progressor patients.
These statistics highlight the need to supplement pathology with an additional risk stratification tool to examine the complex subcellular changes that precede the findings that a pathologist can see under a microscope. Pooled analysis shows TissueCypher to be the strongest predictor of progression to esophageal cancer that we have today.
T he left panel shows the results from real-world pathologist risk stratification, demonstrating little to no ability to differentiate which Barrett's patients are at risk of progression. In a pooled analysis of 5 clinical validation studies, including almost 700 patients, the risk stratification performance for traditional pathology was poor. The rate of progression over time was essentially the same between patients diagnosed with non-dysplastic and low-grade disease.
On this right panel, for the same patients, you can see that TissueCypher identified more than 60% of progressors, many of whom were non-dysplastic.
Independent of the histologic diagnosis, the rate of progression for high-risk TissueCypher score was 7.7 times higher compared to a low risk score. Taking a step back, it's important to note that this test does two important things. The first is that it identifies high-risk Barrett's patients that either harbor prevalent high-grade dysplasia or cancer or are at increased risk of progression to esophageal cancer within 5 years, enabling earlier intervention or more frequent surveillance.
Secondly, it identifies low-risk patients who are unlikely to progress, potentially avoiding unnecessary interventions. As a gastroenterologist myself, what is compelling about both of these endpoints is that they are both indeed clinically actionable. What does a high-risk score mean in practical terms? Based on published literature, patients scoring high risk should be reviewed closely for consideration of an individualized and intensified care pathway.
This is because they are at increased risk of progression, even if they are non-dysplastic. After receiving a high-risk score, non-dysplastic patients are at 18-fold higher risk of progression. Low-grade patients are at greater than 6 times higher risk. Patients harboring prevalent disease are 46 times more likely to return a high-risk score. In other words, if a patient happens to be prevalent, harboring prevalent disease that was missed on endoscopy, the standard of care would fail you.
Additionally, if a patient is not harboring prevalent disease but scores high risk, they are at a 9.4 times higher risk of progression in the next 1 to 5 years. Patients scoring low risk are at lower risk of progression than is indicated by histology alone.
After receiving a low-risk score, non-dysplastic patients are 3.2 times less likely to progress, and low-grade patients are 2.5 times less likely to progress compared to histologic assessment alone. When combined with a high-quality endoscopy, this enables the clinician to feel much more confident in aligning their care with GI Society guideline recommendations.
A significant recognition by a GI society was TissueCypher's inclusion in the American Gastroenterological Association in its clinical practice update in 2022. This practice update focused on new technology and innovation in Barrett's esophagus and was co-authored by many of the luminaries in the field of esophagology.
Best practice advice statement number 9 recognized TissueCypher as a tool that may be used by physicians to stratify non-dysplastic patients. They cited published evidence that indicates a high-risk non-dysplastic patient progresses at a rate of 6.9%, which is similar to pathology-diagnosed low-grade dysplasia.
The update also suggested a care pathway that indicates the use of TissueCypher for risk stratification of newly diagnosed Barrett's patients, as well as patients under surveillance. To understand how TissueCypher fits into one's everyday GI practice, it's very useful to look at current guideline recommendations in the context of 5-year risk of progression.
For non-dysplastic patients, which are the vast majority of cases, the risk of progression is 3.15% over 5 years. Based on this risk, guidelines recommend a 3- to 5-year surveillance interval.
For persistent indefinite patients with controlled reflux, the risk is 7.5%, and guidelines suggest a one-year surveillance interval. Low-grade patients have an 8.5% increased risk, which is enough to consider endoscopic eradication therapy or short-interval 6- to 12-month surveillance based on practice patterns.
A high-risk TissueCypher score in both non-dysplastic and low-grade patients increases the cumulative risk of progression, shown by the red bars, into a range where endoscopic eradication therapy or increase in surveillance could be considered based on our learnings from low-grade dysplasia.
Conversely, for a low-risk score, a pooled analysis shows that the risk drops below non-dysplastic Barrett's esophagus and provides support for considering a less intensive treatment plan in line with clinical guidelines. A final message on incorporating TissueCypher into clinical practice. The test is not meant to replace pathology review.
TissueCypher can and should integrate with guidelines and be used in conjunction with histology to inform precision medicine decisions. Based on the available literature, TissueCypher can be considered for use in non-dysplastic, indefinite, and low-grade histology. A low-risk score in concert with a high-quality endoscopy supports moving follow-up exams of non-dysplastic patients out to 3-5 years.
For indefinite and low-grade patients, a low-risk score may be interpreted as a chance to avoid unnecessary intervention. A high-risk score considerably increases risk of progression, and data would support a more intensive treatment plan. First, it's important to rule out missed prevalent high-grade dysplasia or esophageal cancer.
For non-dysplastic patients with a high-risk score, it is therefore reasonable to recommend a short interval surveillance to 1 year or consider endoscopic eradication therapy based on local practice.
For indefinite or low-grade patients, a high-risk score should be taken very seriously, with consideration given to endoscopic eradication therapy and short-term follow-up. Now a video from one of our key academic collaborators, Dr. Srinadh Komanduri. Dr. Komanduri is an internationally recognized expert in advanced endoscopy and Barrett's esophagus.
He serves as the associate chief of GI and director of endoscopy at Northwestern Medicine. Dr. Komanduri co-chairs the AGA Center for GI Innovation and Technology and was one of the lead authors on the clinical practice update that I mentioned earlier. With that, I'll turn it over to him.
The purpose of this was we have new innovation that's been very validated and is ready for prime time, so we need to create a statement that allows it to be used in practice. We gotta get more clinicians using the assay in their practice to better understand where the value is gonna be. I think the clinical practice update has recognized the investment and the time put into this innovation.
It's recognized the data and the quality of the data to open that door finally for us to actually utilize this clinically and identify where it best fits in the long term.
Hello. My name is Bob Cook, Senior Vice President of Research and Development with Castle Biosciences. Today, I'll be introducing the latest product in Castle's portfolio, IDgenetix. Today's healthcare system and the expectations of patients of all ages are being guided by customer experiences of businesses like Amazon and Tesla. Bringing that same level of customer care and focus to healthcare means that there are new standards for therapies and the benefits that they provide to patients.
The expectation is that a patient will receive the right therapy the first time they're prescribed a drug, that that drug will achieve a fast response with fewer side effects and/or adverse events, and that the drug will have low out-of-pocket costs for the patient.
When it comes to the field of mental health and the 50 million patients experiencing mental health illness in the United States today, a new approach to medication selection is needed to meet the expectations of patients and their providers. Medication selection for patients with mental illness has been challenging for a number of reasons.
The first of those is the inadequacy of therapy response for the majority of patients. Using the current standard of care, less than half of patients with major depressive disorder achieve an adequate response to first-line treatment. A second factor contributing to the challenges of treating mental illness is that nearly three of four patients did not achieve a remission of their symptoms.
Finally, there's a high prevalence of adverse drug events and increasing rates of discontinuation with repeated medical trials. Thus, the process of finding an effective therapy for patients diagnosed with mental illness can take years and still not achieve an adequate patient response. IDgenetix is setting a new standard for pharmacogenomic testing to provide the most comprehensive assessment of patient characteristics to guide treatment decisions.
Next-generation sequencing, next-generation PGX testing with IDgenetix provides a bioinformatics platform that combines drug-gene and drug-drug interactions, along with lifestyle factors, into a clinically actionable report that provides a precision medicine solution that can overcome trial-and-error prescribing of drugs. The test has been clinically validated in a randomized controlled trial.
It's been published in the Journal of Psychiatric Research, and the test has demonstrated unrivaled efficacy and provides information about 26 drug classes, 15 genes, 127 medications, and 10 mental health and pain conditions. With less than 1 minute required to collect tissue sample and a 3-5-day turnaround time, the IDgenetix test is fast and easy to use. IDgenetix is supported by a large multi-center randomized controlled clinical study of 685 patients with depression and/or anxiety.
The study was published in the peer-reviewed Journal of Psychiatric Research. Data from this study demonstrated that the use of IDgenetix test enhanced therapeutic efficacy. At week 12, patients with severe depression in the IDgenetix guided group had twice the response rate compared to that of patients in the standard of care control group. Response was defined as 50% or greater reduction in Hamilton score, the most widely used clinician-administered depression assessment scale.
Remission rates at week 12 are shown on the right side of the slide. Patients who were in the IDgenetix arm of the study were more than 2.5 times as likely to achieve remission compared with patients in the standard of care control group. This study demonstrates clinical utility of IDgenetix for depression and anxiety.
Based on the results of this study, which was conducted at 20 independent sites across the US, IDgenetix improves outcomes for patients with both depression and anxiety in a diverse set of relevant clinical settings. IDgenetix offers pharmacogenomic testing that provides important enhancements beyond the tools that are available for physicians today. IDgenetix is a multi-gene test that has published randomized controlled clinical trial and robust clinical utility data.
IDgenetix is covered by Medicare and is supported by published data demonstrating use in patients with dual major depressive disorder and anxiety diagnoses. The characteristic that separates IDgenetix from other testing is that it goes beyond drug-gene interactions.
The IDgenetix integrated bioinformatics platform also includes drug-drug interactions and lifestyle factors to provide a more comprehensive test to guide therapy selection and help improve medication response and remission rates.
The IDgenetix test report is designed with two simple classifications that clinicians can use to guide medication choices. The green use as directed classification requires no adjustments. The yellow use with caution and/or increased monitoring classification indicates a need for additional attention by the physician. This slide presents two patient scenarios. On the left is patient A, who's currently taking an antidepressant and ADHD medication.
In considering further medications, these medications and any lifestyle factors would be listed on the test requisition form. After the test is completed, the report indicates no genetic variants or metabolic interactions that may alter his response to these medications. All medications are classified as use as directed, as shown in green, meaning no additional adjustments would be necessary by the physician.
Patient B on the right side of this slide is currently taking anti-anxiety medication and a pain reliever. She's also a smoker. After running IDgenetix, her personalized profile alerts her physician that there are medications classified as use with caution and/or increased monitoring.
This means that the provider should use increased caution when prescribing new medications, make a dose adjustment or change medications potentially because of drug-gene or drug-drug interactions, or because of the impact that smoking has on gene function. IDgenetix is innovating the process of treatment selection for the many, many patients suffering from mental illness.
Patients and doctors now have the opportunity to move away from the trial-and-error based approach that doesn't include genomic testing.
Into a new world where drug-gene and drug-drug interactions can be combined with lifestyle factors to provide personalized information that guides appropriate treatment selection to improve response and remission rates for those patients who are suffering from mental health illnesses. I'd like to thank you for your time and attention in joining us today.
Now I look forward to discussing our ESG initiatives and our financial overview. Our culture and operations have been built on ESG principles. Those principles help to mitigate risk and create opportunities for our company to meaningfully transform patient management in the areas of dermatology, gastroenterology, and mental health. We also believe that the breadth of ESG principles builds employee engagement and improves employee retention and satisfaction.
In the fourth quarter of 2021, we formalized our ESG program by releasing our inaugural ESG report, giving increased transparency for both ourselves internally as well as our external stakeholders. The launch of this inaugural report marked an important milestone in our journey, demonstrating our commitment to moving Castle forward and progressing our ESG goals.
Additionally, it reinforces our commitment to improve the lives of patients, positively impact our communities, and ensure that Castle remains a great place for our valued employees to learn and grow. Review and oversight of our ESG program resides with the audit committee of the board. The audit committee periodically provides reports to the board of directors on ESG matters.
Additionally, we have established an internal management committee comprised of a cross-functional set of representatives, including leaders from marketing, finance, human resources, and operations to develop strategy and execute on these matters. On a periodic basis, this management committee reports to the audit committee.
We recognize the importance of aligning our business practices to robust ESG principles, and we are committed to identifying and executing opportunities for improvement. We also understand that the development of appropriate governance and oversight structures is a critical component of those efforts.
Some areas in which we intend to focus in the future include environmental policy, environmental metrics, DEI mission statement, DEI metrics and action plan, and vendor code of conduct and supplier standards. Our philosophy of corporate responsibility and giving focuses on patient care, education, and community engagement.
Most of our education and most of our current educational giving is focused around helping to improve the education of patients who are diagnosed or who may be at risk of skin cancer. This improves things that are peripheral to Castle. One of those is to promote skin prevention, skin cancer prevention strategies.
If we can help people understand the risks of developing melanoma or squamous cell carcinoma and be proactive and preventative in reducing that risk, that's a terrific outcome for patients. Finally, on community engagement, we have the same philosophy and approach.
We endeavor to help improve early detection, improve prevention, and improve management strategies so that patients who are diagnosed or perhaps diagnosed earlier and caught earlier will have improved outcomes for their health. Turning now to a financial overview of Castle. Our financial focus is built on five core pillars.
The first is the continued driving of robust test volume growth. Initiatives that support that, for example, recently include the expansion of the TissueCypher sales force, as well as the carve-out of a small, dedicated group to focus on MyPath Melanoma and DiffDx-Melanoma. We believe that we'll see the benefits of that beginning in the fourth quarter this year and accelerating into 2023.
In addition, we're focused on our industry-leading gross margins, and we expect to continue to generate industry-leading margins, and the key to that effort will be continuing to aggressively work to reduce the number of unreimbursed test reports.
The cost of goods of those reports, as well as the associated marketing expense, is already reflected in our P&L. As we convert unreimbursed tests to reimbursed tests, the associated revenue drops all the way down the P&L. We intend to maintain a strong balance sheet. We finished June thirtieth, 2022 with $273 million total cash, and our capital structure is debt-free. We expect our current cash position to meet our long, our near and midterm capital allocation plans.
Finally, as we've said before, we expect to achieve operating cash flow neutrality by 2025. Our second quarter financial results continue the strong performance that we saw in 2021 and the beginning of 2022. Total test reports in Q2 were up 57% from a year ago to 11,034 reports.
This was driven by strong gains in the Derm franchise, which was up 44%, as well as contributions from TissueCypher and IDgenetix, which both began contributing sales in the first half of this year. Revenues were up 53% to $34.8 million, and adjusted revenue, which excludes the effect of revenue adjustments related to tests delivered in prior periods, were up 49% to $34.3 million.
Our operating cash flow for the quarter was a $9 million use of cash, as was our adjusted operating cash flow. As I mentioned earlier, we ended the quarter with a cash balance of $273 million.
Over the next three years, we plan to achieve revenue growth of 25%-35% per year, with 2025 revenue expected in the range of $255 million-$330 million. Our adjusted gross margins will trend back up, with 2025 having margins in the low-to-mid 80% range. Operating expenses should stabilize around 75%-85% of our revenue by 2025. By 2025, we expect to be operating cash flow positive and do not expect to have a need to raise capital to reach that point.
This graphic demonstrates the penetration of our current commercial tests, and importantly, also demonstrates the long runway of growth we have ahead. We are frequently asked what is the maximum level of penetration for a Castle test?
We think that each of our tests should reach at least 50%-60% penetration, and Castle has a history of success here. Our DecisionDx-UM test is standard of care, and we believe we test 80%-85% of patients diagnosed with uveal melanoma. Other areas of cancer screening have seen penetration levels above the 50%-60% level as well. Our flagship test, DecisionDx-Melanoma, has grown very strongly, and we believe the market penetration is roughly 20% of patient diagnosis.
We are also encouraged with the adoption of our DecisionDx-SCC test, the first to market test for SCC patients with one or more high risk factors.
More than 2,100 clinicians have ordered DecisionDx-SCC since it was first launched, and those 2,100 clinicians have ordered almost 8,000 tests for patients with one or more high-risk features. The bulk of this test ordering has come from Castle's core group of dermatologists who are already ordering DecisionDx-Melanoma, as well as Mohs surgeons and dermatology practice extenders.
Interestingly, about 10% of our SCC orders are being driven by a broad group of multidisciplinary specialty clinicians, including radiation oncologists, ENTs, head and neck surgeons, and plastic surgeons. As we drill down into the performance of DecisionDx-Melanoma, we note that our growth has taken penetration from about 2% to about 18% at the end of last year.
The test report volume CAGR during that period was nearly 39% as volumes grew from 2,858 tests to 20,328 tests. We continue to believe that our strong body of evidence supports adoption, and we have more than 35 peer-reviewed publications supporting the clinical use of DecisionDx-Melanoma. Earlier, Dr. Moody discussed our NCI-SEER collaboration data. The publication of that data will be a powerful tool for educating physicians about the benefit of DecisionDx-Melanoma.
This year we reached a very important milestone when orders for DecisionDx-Melanoma surpassed the 100,000 level. As a reminder, DecisionDx-Melanoma is the only GEP test for cutaneous melanoma that informs two clinical treatment questions in the management of melanoma.
The first being a patient's individual risk of sentinel lymph node positivity, and the second being a patient's personal risk of recurrence and or metastasis. Starting from 2018 and running through 2021, our revenue CAGR was 60%, ending 2021 with full year revenue of $94.1 million. Adjusted gross profit had a CAGR of almost 64% over the same time period, ending at $74.9 million for 2021. Our adjusted gross margin line has been very stable as well.
Over the last two years, we've been very active in adding new product lines and verticals, both organically and through acquisitions. For example, in the second half of 2020, despite the impacts of COVID disruptions and shutdowns, we had the commercial launch of both DecisionDx-SCC and DiffDx-Melanoma.
In 2021, we acquired MyPath Melanoma and Cernostics, which brought us the TissueCypher test, and most recently we acquired AltheaDx. Our expanded test offering has also driven increased expenses in categories as well. As you can see, R&D and SG&A in particular have risen as we've added to our product portfolio.
As we mature our newly launched tests, we believe revenue gains, process improvements, and operating leverage will ultimately yield improved profitability over the next three years. As I noted at the end of the second quarter, we had $273 million of cash and 0 debt. Our balance sheet positions us well for continued growth and value creation.
Our priorities for capital spending fall into three main categories. First, and most importantly, is supporting our current products and continuing to advance our pipeline products.
Core to this priority is the building on our R&D efforts to solidify our base of evidentiary support, as well as maintaining the momentum and pace of developments of products in our pipeline. Obviously, this evidence is critical to further test adoption and our future growth.
Our second focus is on commercial optimization, which means adjustments to our commercial team as well as facility enhancements and other workflow adjustments and improvements intended to boost efficiency. These efforts will be ongoing as we look to balance a growing number of tests and test report volumes in our portfolio with a need for greater productivity and operational efficiency.
Finally, we will keep an eye out for possible bolt-on acquisitions to our current three franchises. Although I think from a strategic perspective, this is well down the list of capital priorities.
We are excited about what we have in our portfolio right now, and that's where we're going to focus our capital, pushing those sales and advancing those programs. By being disciplined in our capital spending and allocation, we will be it will help us reach our goal of being net operating cash flow neutral by 2025.
As our business matures, we expect that growing test volumes, efficiency gains, and measured adjustments to our cost structure will lead to attractive margins and a profitable model. With this graphic, our P&L is portrayed as a waterfall chart in a common size format, displaying the major line items as a percentage of revenue using a base figure of 100%. We expect that at maturity, COGS will stabilize at roughly 20% of revenue, leaving us with gross margins of approximately 80%.
SG&A will remain our single largest operating expense, but should be reduced to approximately 40% of revenue, and R&D expense should be approximately 20% of revenue, leaving us with operating margins of approximately 20%. Ultimately, we believe that as revenue rises, operating leverage will work in our favor, and combined with appropriate expense rationalizations where needed, bring us to a right-sized cost structure that can yield a sustainable and profitable model.
Let me summarize some of the key financial highlights for Castle. As you've seen based on our financial results so far in 2022, our tests have demonstrated clinical validation and utility that physicians and other healthcare providers value and appreciate.
This, combined with our ongoing commercialization efforts to educate, inform, and market to new and existing customers, has driven greater adoption of our tests as clinicians have sought to make more informed treatment plan decisions. The rise in test volumes, along with our ongoing efforts to push reimbursement among third-party payers, will drive revenue gains that put us in a position to support and grow our margins.
We have made several noteworthy product introductions and focused growth investments over the last couple of years to help ensure Castle's long-term growth prospects. While this has raised our cost structure, we expect to optimize these investments and achieve improved workflows and greater efficiency as our operations mature. With that in mind, our expectation is that we will be operating cash flow positive by 2025.
Keep in mind that we have a strong debt-free balance sheet, which gives us sufficient capital to reach that goal. We also remain focused on driving improvements in our newer acquisitions with an emphasis on more measured capital spending as we digest these new opportunities.
This is an exciting time for all Castle stakeholders, not just in terms of our financial condition, but also opportunities to improve the care of more patients each year. With that, I'll turn it back over to Derek for a quick summary.
Thank you, Frank. As I wrap up our Investor Day discussion, let me review a couple of summary slides, and we can then move on to Q&A. Upcoming milestones. Please keep in mind that these are forward-looking expectations as things stand now, so circumstances can shift as we make our way through the balance of 2022 and into 2023.
One of the most anticipated milestones would be the expected publication of the peer-reviewed study by Castle Biosciences and our collaboration with the National Cancer Institute using the matched SEER Cancer Registry data. We expect this to be published towards the end of this year or certainly in the first half of 2023.
Though initially showed in poster form in April 2022, this poster form presentation did have an initial impact in terms of establishing melanoma's credibility as well as driving physician adoption volumes and test volumes. Once the collaborative study is published in manuscript form, we believe this will serve as an additional powerful catalyst in further driving greater product awareness, greater product value, and thus spurring further adoption and market penetration.
Another milestone will be the finalization of the Palmetto GBA/Noridian LCD for our DiffDx-Melanoma test. Following the draft LCD posted in June of this year, we're hoping for finalization by the end of 2Q 2023. If this comes to pass, we'll have Medicare coverage for both our MyPath Melanoma test and our DiffDx-Melanoma test.
As we have also discussed, we expect the Palmetto draft LCD for our squamous cell carcinoma test, as this will be a big driver of test adoption and market penetration. Right now, we are hoping that this comes in the second half of 2022, with finalization later in 2023. The process is rather opaque and therefore timing is quite speculative at this time and subject to change.
In 2Q 2023, we expect our expanded gastroenterology sales force and our expanded diagnostic GEP sales force and/or commercial teams will really begin to start hitting their stride and ramp up in productivity that we'd expect to see throughout the rest of 2023.
We also expect to close the San Diego laboratory and integrate all of the operations of IDgenetix testing into our Phoenix laboratory by year-end. As you can see here how we expect our franchises to contribute to revenue in the near term, midterm, and long term. Note that I have left DecisionDx-UM melanoma off this slide due to what we believe is a 80%-85% market penetration, so it will not be a contributor to growth going forward.
Currently, we have Medicare coverage for DecisionDx-Melanoma, DecisionDx-SCC, MyPath Melanoma, TissueCypher and IDgenetix, which as you may recall, was part of our acquisition strategy to target tests with existing coverage. We have six proprietary tests with Medicare coverage and varying levels of commercial coverage as of today. That's a position of strength.
As we think about Castle's long-term potential, we expect our two acquired franchises to be material contributors in addition to continued organic growth of our dermatology business. We believe that the pipeline programs for all three franchises will begin to generate additional tests in their respective markets in 2025 and beyond. We have confidence in our ability to deliver on our 2025 revenue target of between $255 million and $330 million.
As I close out here, I'd like to go ahead and remind you about what we're in business for, and that is really in serving patients. I'd like to revisit some of the data that Dr. Moody shared earlier. As doctors and patients continue to desire more information when they make shared decisions regarding their treatment plans.
This slide represents two studies that we conducted in collaboration with the Melanoma Research Foundation. Two studies of patients diagnosed with either cutaneous melanoma and UM melanoma. As you can see, the data for both forms of melanoma align quite nicely. 90% of the patients who participated in the survey reported wanting prognostic information at the time of their diagnosis.
This is not necessarily an uncommon expectation as we see the same kinds of desire in patients diagnosed with, say, breast cancer. Patients in this decade are informed. They want to be more informed. They want to be part of the shared decision-making for their cancer. What we saw here is that patients who were diagnosed with either cutaneous melanoma or UM melanoma, nine out of ten patients want information about their diagnosis at the time of diagnosis.
They can help understand what they're facing and what treatment plan decisions they want to go forward on. That's fantastic.
As from Castle Biosciences' perspective, to also have 92% and 99% of these patients also say, "We found the information useful in DecisionDx-Melanoma or useful in DecisionDx-UM in helping us to understand that the testing was useful," that they gained value from the test result, and ultimately, at the end of the day, that would improve the shared decision-making responsibilities between their physicians and themselves.
That is what we're in business for. To finalize here, as we end today, I wanna thank you for joining us. Thank you for your continued interest in Castle.
We believe our guideposts for the future offers us a focused path ahead and further positions us to impact patient care and value creation in the future. We will now open up for Q&A.
Thank you for joining today. This is Camilla Zuckero, Vice President of Investor Relations and Corporate Affairs. We will now begin the Q&A portion of the day. If you would like to submit a question for Derek, Frank or Doctors Monroe or Goldberg, please use the Ask a Question button on the top right-hand side of your screen. We may not have time for all to answer all of your questions live today, but I will have your information here.
The first question is from Puneet Souda from SVB Leerink. Should we expect Castle to pursue other verticals as part of strategic opportunities?
Hi, Puneet. This is Derek Maetzold. That's an excellent question. I think as part of our strategic opportunity, certainly going forward in the near term, what we would like to do is to continue to go ahead and flesh out the sales bag for these three verticals that we're investing in: dermatology, gastroenterology, and the mental health segment.
The next question here is from Thomas Flaten from Lake Street. Can you provide an update on the current status and near-term plans for the mental health sales team?
I'll take that one too. Yes, I can. When we acquired or closed the transaction in late April this year, there were approximately, I think, 18 outside sales territories. We plan to maintain roughly that same level of sales territories going into the winter and probably the first quarter of next year. As we complete the integration of this acquisition, we will make planned, thoughtful investments in terms of expanding the field force over the perhaps end of 2023, going into 2024.
In the very, very near term, the main goal here is to really manage messaging, manage training, get ourselves organized, integrated as a single company rather than two companies, and then we can go ahead and place our expansion opportunities going forward next year.
The next question is also from Thomas. What is the status of the rollout of the MyPath/DiffDx-Melanoma sales team?
Excellent question again, Thomas. We have completed a hiring of all of the territories for the expansion into the, I would call this, the dermatopathology focus sales team. Along those same lines, we also completed, I believe it was early last week, the hiring of the additional gastroenterology sales representatives. Those two groups will be managing their way through training in the next month or two, and we expect them to achieve reasonable productivity going into 2023.
All right, next question, coming back to Thomas is, will you continue to run MyPath first should DiffDx upon finalization of LCD? And what other considerations should we be aware of that might change this process?
Yeah. Let me back up here so the audience has a bit more of a background here on top of Dr. Goldberg's earlier comments. We offer, as you know, two tests to assist a dermatopathologist and a dermatologist in getting to a more clear diagnosis. Is this lesion that's sitting in that gray area of diagnostic clarity looking more like a malignant melanoma or more like a benign lesion? These two tests have similar sensitivity and specificity.
Each one of them appear to have some unique benefits from a patient care perspective. Our decision upon the acquisition of MyPath Melanoma in May of 2021 was to, in the absence of an ordering clinician's request, that we would go ahead and process the MyPath Melanoma test first.
Why is that predominantly based upon the fact that the overwhelming evidence that was published, when you compare MyPath Melanoma to DiffDx-Melanoma, just due to years in the marketplace, was in favor of MyPath Melanoma, and we believe that it was in our best customers' interest, our best patient care interest, to offer the test with the most data first.
Should we end up having an intermediate score come back on the MyPath Melanoma test or a failure to have the test processed well or report an actionable test result, then we would go ahead and offer up running the DiffDx-Melanoma as a secondary or backup test, I would say. Going forward, we would expect a few things to happen.
We are, as you may be expecting, running a large comparative study in the background of clinical order processing.
Our expectation is that study may finalize in the course of 2023, and should it change a direction in better patient care, then we may go ahead and just alert our customer audience or our customer base saying, "Hey, by the way, you can still order whichever test you feel most comfortable with.
But based upon this newer data, we might recommend that you still continue with MyPath first." We might find that the data actually points to in certain patients or all patients that we'd run DiffDx-Melanoma first and offer you MyPath Melanoma on the ones that produce the intermediate results.
I think we would say, we'll let the data mature and help us lead that direction which way we should go from a best patient care perspective, and that should happen, I think, Dr. Goldberg, probably in later 2023. But do you wanna add commentary there?
No. I just think that commitment to billing just one test is the other thing to emphasize, that Castle Biosciences pursues a workflow such that even in a situation where two gene expression profile tests are run, the result that's reported to the patient would lead to only one bill will be returned to the beneficiary.
That's excellent clarity there. Yeah.
Okay, next question from Mason Carrico from Stephens Inc. Historically, I think your 65 rep dermatology team had focused the majority of their time on DecisionDx-Melanoma. With DecisionDx-SCC reimbursement now in place, could you provide an update on how you intend to focus this team in terms of melanoma versus SCC?
That's me again. Okay. Thank you, Mason. To kind of rewind a bit over the last month or so, we made the decision going into July that we were going to establish or reestablish a small team focused on dermatopathology so that we could have an appropriate share of voice targeting MyPath Melanoma and DiffDx towards the dermatopathology customer base.
In doing that ends up freeing up the 65 plus or minus person sales team focused on dermatology to really focus on predominantly the cutaneous melanoma test and the squamous cell carcinoma test. Going into the fourth quarter here, we are adjusting our focus from before being sort of 80% DecisionDx-Melanoma, roughly 10% DecisionDx-SCC, and roughly 10% MyPath Melanoma/DiffDx-Melanoma to being approximately 70/30, 60/40, DecisionDx-Melanoma/DecisionDx-SCC.
That's a change we made effective in the middle of this quarter, so we would expect to see some adjustments or some change in velocity of growth, I think for our squamous cell carcinoma test probably in the first quarter of 2023. Frank, do you wanna add any clarity?
No, I think that's good.
All right, next question from Tom Peterson with Baird. Can you explain in a little more detail the range of the $255 million-$330 million 2025 revenue target? What gets you to the high end of that outlook?
Frank?
Yeah. Thanks, Tom. It's of course just,
The variability in that growth rate, but I think the drivers that would push us to one end of the range or the other, you know, are primarily related to our earlier stage in terms of marketing products. We've got a fair amount of experience with melanoma at this point. I think we have a high level of comfort and confidence in our modeling there. We're newer in squamous cell, we're newer in the launch there.
Of course, as well as with the differentials product and with the newly acquired products, we're still early in those launches. The biggest variable, I think, is the impact of those less mature products.
Okay, another question here from Thomas. From slide 94, with the lesser prioritization of tuck-ins, how long do you think it will be viable to carry only a single product in each of the GI and mental health verticals?
Right. Yeah. Thanks, Thomas. We certainly believe that the TAMs of those two opportunities can support a single product. We, as I think we said before, we'll scale the sales organizations appropriately as volumes grow. There’s an awfully big pool there to go after. We're comfortable those products alone could support a thoughtfully sized sales force.
All right, what looks like here, our last question. How should we model DecisionDx-SCC in 2023, given the current uncertainty surrounding CMS reimbursement?
Right. Yeah. It's an area of some uncertainty, as you noted. I think that we've got a high level of confidence in the clinical utility of that test, and we've got a high level of confidence in the evidence we've built. We do believe that while we can't predict the timing of the various processes that impact that reimbursement, we think it's a win rather than an if, in terms of stability and longer term line of sight on reimbursement.
The market for squamous is a larger patient population. It is a potentially even more impactful clinically than our melanoma, which as you know, has tremendous clinical impact.
It's a matter of when we get that settled out. I think that if you'd like to take 2023 and model a base with SCC as an upside, I'm okay with that, because we've got great growth underlying that in either case. We're hopeful by the end of the year we can have a little bit more visibility and continue to increase the transparency and the ability for you to confidently model that.
That looks like that's all we have time for today. We're gonna close here. Again, thank you for joining.